This paper provides a detailed quantitative description and analysis of the courtship and reproductive actions of Trichogaster leeri. Principal objectives are to present a description which will be useful in future comparative (and phylogenetic) studies, to determine how well detailed quantitative analysis provides insights into motivational mechanisms in these fish, and to see how well such data can be interpreted in terms of conventional ethological theories. Five pairs of mature fish provided complete records of fifteen spawnings (referred to below as sequences) ranging in duration from 2.7 to 5.5 hours. Reproductive and courtship activities in captive T. leeri and other anabantoid fishes occur in bouts of varying length which may be categorized on the basis of behavioral components present. While most bouts contain behaviors involving both fish, some bouts occur in which one fish (usually the female) shows no visible response to the action of the other. Examination of the organization of these bouts in the spawning sequence, the nature of activities in the different types of bouts, and the relationships between completeness of the bout, sex of the fish initiating the bout, and duration of intervals between bouts provide information about some of the factors influencing courtship and spawning. Bouts that are initiated by females are much more likely to be successful (proceed to clasping or spawning) than those initiated by males. Although males initiated over half of the bouts, spawning occurred in 88 female-initiated bouts and in only 10 male-initiated bouts. Thus we suggest that while male overt activity is often prominent several days to weeks before and after spawning, it is actually less critical in determining the outcome of sexual bouts than are female activities. Highly motivated males had little success in stimulating refractory females to successful spawning, whereas the converse appeared commonplace. Data on duration of intervals between bouts shows that females generally remained away from the nest and male longer after more complete bouts than after less complete (and typically shorter) bouts. Thus it seems unlikely that sexual and agonistic interplay during the bouts stimulated the female directly (arousal function). Furthermore, male-initiated bouts were found to be more successful the greater the duration of the interval preceding his approach. Although part of this apparent refractoriness is undoubtedly attributable to muscle fatigue or a similar phenomenon, performance of more complex actions in the more complete bouts may have some kind of central satiating (or drive-reducing) function. In the next section, variations in the frequency and duration of female approach, male leading-to-the-nest, male lateral spread, male circling, female courtship butting, male clasp, swimming inhibition, male butting and biting, male chasing, female appeasement, and total bout durations are presented in tables as they occurred in the 15 different types of bouts. Possible causes for these variations are discussed and the use of bout statistics in evaluating motivation levels is examined. For example, detailed examination of data on male butting and biting and male chasing suggest that male aggression only bouts may actually represent a rather lower degree of absolute aggressiveness in males than that which occurs in many bouts that succeed to more complex levels. Female sexual signals or responses may enhance aggression as part of a general increase in excitement, or partial female responses may lead to thwarting and resultant increased aggression. Much data is presented to support the contention that male courtship only bouts are least associated with aggressive motivation and are a good indicator of sexual motivation in the male. Furthermore, we suggest that the frequency of these two types of bouts in a sequence may be a fairly accurate reflection (if such is actually possible) of the relative strengths of sexual vs. aggressive motivation in a given sequence, because these two bout types occur in the absence of any measurable female responses, thus eliminating that critical variable from the list of factors which might influence male behavior. Evidence also is cited to support the contentions that female flight tendency is lower in female-initiated bouts and that male aggressiveness tends to be slightly higher in many (if not most) male-initiated bouts than in female-initiated bouts. There is also an indication that the nature of actions occurring in bouts have less influence on the activities of males in subsequent bouts than on those of females. In the next section, data on 10 of 11 activities discussed above in terms of their appearance in all bouts of a certain type (across the 15 sequences) are presented as they occur in each sequence, regardless of the bout type in which they occurred. The sequences are ranked (Table 18) from those in which the males were judged most aggressive (measured by frequency of male butts and bites) to those judged least aggressive. Aggression analysis shows that another potential measure of aggressiveness, male chasing frequency and duration, shows no direct correlation wih butting frequency, except perhaps a slightly negative one. The reasons for this are discussed in detail. The choice of male butting rate as the most nearly accurate indicator of male aggressiveness is justified on the basis of its agreement with other parameters such as female appeasement duration and circling duration (previously and subsequently shown to be associated with several indicators of aggression). High-aggression sequences tend to be those in which female responsiveness (or sexual initiative) is minimal. If male aggressiveness is determined exclusively internally, it is likely that female self-protection butting would occur more commonly than it apparently does in such sequences. It seems probable, therefore, that female behaviors may markedly effect motivation levels in the male. Since highly-aggressive males also seem to exhibit much leading-to-the-nest and other indicators of sexual motivation we suggest that factors causing increased approach tendency, aggression, and courtship in males are not entirely independent of one another. In the sexual motivation analysis, data are discussed which indicate that high clasp frequency and low clasp durations in a sequence are indicative of high female sexual motivation. Such sequences also have a high percentage of female-initiated bouts. Conventional analysis of male leading-to-nest, lateral spread display, and circling can be interpreted to support the hypothesis that the three behaviors represent three conflict levels between aggressive and sexual systems, with leading dominated by sexual factors, lateral spread indicating near equilibrium, and long-duration circling dominated by aggressive factors. Female butting is a sexual signal indicating readiness for further sexual activity and there is a strong indication that it is used also as a self-protection device in the absence of female readiness to spawn. Examination of spawnings of the same pairs in a temporal sequence shows that female butting rate increased in later sequences in all cases, suggesting the importance of experiential factors in the ultimate regulation of this activity. In the section on organization of motivating mechanisms, one pattern, the lateral spread display, was selected for more detailed analysis in order to evaluate the utility of generally-accepted ideas on conflict theory and attack-escape motivation. We are forced to reject the theory that lateral spread can be strictly interpreted in terms of a certain A/F ratio. Rather, the idea that fin spreading need not be under the direct control of motivational systems (at least initially) is expanded to encompass the possibility that learning processes may be intimately involved in the development of the lateral spread. The senior author presently is testing this typothesis in Trichogaster trichopterus and Macropodus opercularis. Our final conclusion is that alternate hypotheses appear to be as useful as conventional attack-escape theory in interpreting this extremely common behavior. In the final section we have tried to see whether or not our data can provide some insight into the nature of phenomena such as positive or negative arousal in T. leeri. Although we have shown in an earlier section that longer, more successful bouts were followed by longer inactivity in females, other data indicate that female-initiated bouts are generally followed by female-initiated bouts which are as complex or more complex than the initial bout. The latter pattern (in females) might be interpreted as an indication of arousal. If so, it is possible that performance of courtship and sexual behaviors has both positive and negative effects on arousal. Interestingly, some degree of success in male-initiated bouts tends to increase the probability of further male courtship in subsequent bouts. It thus appears that performance and success in actions initiated by fish of each sex has reinforcing properties only on fish of that sex. Although virtually no clues exist as to the nature of the factors that terminate spawning in the female, some of the factors affecting the male apparently are accessible. Data on male approach, aggressiveness, and courting actions throughout a spawning cycle suggest that males are more prone to courtship activity after the terminal spawning bout than at any other time. Since there is no clear evidence of enhanced male courting tendency after other spawnings, the effect must be due to some cumulative process. Failure of the female to respond to male courting gradually leads to increased male aggressiveness, which a few hours after spawning is greater than at any other time.