Accurate estimation of a species' size distribution is a key component of characterizing its ecology, evolution, physiology, and demography. We compared the body size distributions of five Pacific lizards (Carlia ailanpalai, Emoia caeruleocauda, Gehyra mutilata, Hemidactylus frenatus, and Lepidodactylus lugubris) from general herpetological collecting (including visual surveys and glue boards) with those from complete censuses obtained by total removal. All species exhibited the same pattern: general herpetological collecting undersampled juveniles and oversampled mid-sized adults. The bias was greatest for the smallest juveniles and was not statistically evident for newly maturing and very large adults. All of the true size distributions of these continuously breeding species were skewed heavily toward juveniles, more so than the detections obtained from general collecting. A strongly skewed size distribution is not well characterized by the mean or maximum, though those are the statistics routinely reported for species' sizes. We found body mass to be distributed more symmetrically than was snout–vent length, providing an additional rationale for collecting and reporting that size measure.