The effects of injecting prostaglandin F2 alpha (PGF) in dairy cattle within 40 days of calving on the rate of pregnancy at first service and on the number of days open were investigated by meta-analysis, by pooling data from numerous independent trials into single analyses. A total of 4052 cows from 24 trials described in 10 papers were included in the analyses of first service pregnancy rate and 2646 cows from 21 trials described in nine papers in the analyses of the number of days open. The analyses were stratified by trial and by the reproductive health of the cows after calving. Treatment with PGF during this early post partum period had no significant effect on the first service pregnancy rate of cows with a normal or abnormal puerperium. An analysis of the data for the number of days open by effect-size methods showed that a significant percentage (54 per cent) of the treated cows had fewer days open than the untreated cows, and that this difference tended to be greater for cows with an abnormal puerperium (59 per cent of treated cows). The weighted average reduction in days open between the treated and control cows was 2.6 days for trials with abnormal cows, and 3.3 days for trials including normal and abnormal cows. However, the calculation of the fail-safe number (F) indicated that it would require only a few studies with a negative response to PGF to negate this finding. Meta-analysis was a useful technique for evaluating the apparently conflicting results of previous studies.