The influence of microclimate on numbers of third-state larvae of haemonchus contortus and Haemonchus placei in four strata of irrigated Kikuyu pasture was assessed. On 36 different, interspersed days three replicates of pasture samples were collected on three occasions per day from 1990-1992 for larval recovery and the log10 mean counts of the larvae recovered were analysed by use of ANOVA models. Because the ground-surface area from which herbage was collected was standardized, estimated larval counts for the different strata could be compared; this was not possible in our previous studies. For H. contortus, the estimated larval counts in the four strata were predicted by microclimatic air temperature, relative humidity and soil moisture, with the coefficient-of-determination (R2) values ranging from 0.15-0.35. Of these, air temperature had the greatest effect. The same three predictors, together with illumination and wind speed, featured for H. placei, with R2 values of 0.19-0.52. With the exception of wind speed and illumination, which (for H. placei) had the opposite effect, all the microclimatic parameters listed, predicted an increase in numbers of larvae from a lower to an upper strata.