Should a platyopuntia expend all aerolar meristems in flower production, now new cladodes could be produced, and further reproductive effort and vegetative growth would cease. To investigate the trade-off between flower and cladode production, the numbers of flowers, fruits, and cladodes were monitored for 4 years on 30 Opuntia engelmannii Salm-Dyek, plants on Tumamoc Hill, Tucson, Arizona. Plant size controlled the number of flowers initiated each spring. The proportion of flowers that developed (i.e., did not abort) was perhaps determined by December-February rainfall in the months before bloom, with more being developed in the wettest years. Models based on different ratios of initiated cladodes to initiated flowers demonstrated that continued high investment in flowers and fruits would eventually terminate reproduction altogether; therefore periods of high sexual reproduction should alternate with periods of high vegetative growth. In the first 3 years of this study, the ratio of new cladodes to initiated flowers was low, showing a high investment in sexual reproduction. As suggested by the model, the population recouped this investment in the fourth year, when the number of new cladodes was nearly 3 times the 1992-1994 mean, and the number of initiated flowers was only 73% of the 3-year mean.