The abundance of neustonic postlarvae and newly recruited benthic lobsters Homarus americanus was measured for 2 yr along the central coast of Maine, USA. Postlarvae first appeared in significant numbers in late July, rapidly increased in early August to ca 10 to 20 (1000 m²)⁻¹ (0.5 m depth)⁻¹, then gradually declined throughout August and into early September. Late intermolt stages D₁ to D₃ characterized most of the neustonic population, with a shift in dominance to D₃ in the warmer year (1990). Postlarval abundances differed little between years: the seasonally and spatially averaged daily abundances were 5.58 (1000 m²)⁻¹ (0.5 m)⁻¹ ± 0.77 (SE) and 6.76 ± 1.92 in. 1989 and 1990, respectively. In contrast, there was a statistically significant difference in benthic recruitment measured in September. New recruits, lobsters with carapace length ≤ 10 mm, averaged 1.66 ind. m⁻² ± 0.37 (SE) in 1989 and 0.79 ± 0.22 in 1990. Spatial patterns in recruitment density at the sampling sites were positively correlated among years (r₂ = 0.82) and did not appear to be related to distribution patterns of postlarvae. It seems likely that habitat type substantially influenced benthic recruitment patterns. Moreover, local differences in recruitment density of the first cohort were conserved over the 2 seasons of observation, supporting earlier assertions that new recruits are relatively sedentary. We develop a simple diagnostic model of the pelagic-to-benthic transition in lobster recruitment that accounts for the increase in density by 2 orders of magnitude from postlarval to benthic (recruit) stage, and we use this model to suggest plausible rates of postlarval diving and settlement that describe average and sitespecific recruitment in the 2 years.