The use of pulse oximetry in dental procedures involving sedation and children has been encouraged strongly. Reports have indicated that hemoglobin desaturations are not uncommon when a variety of drugs (including nitrous oxide) and dosages are used. The present study investigated the nature of desaturations in relation to patient activity as a function of a drug combination, single drug dose-response design, and placebo conditions. Twenty-two patients (mean age 29.45 months; range 20-37 months) participated in this two-part study. In part 1, 12 patients received a placebo and a combination of chloral hydrate and hydroxyzine in a crossover, double-blind design. In part 2, 10 patients received a placebo and three doses of chloral hydrate in a double-blind, randomized sequence of four visits involving a dose-response design. During each restorative visit, the peripheral oxygen saturation, expired carbon dioxide, heart and respiratory rates, blood pressure, and frontalis electromyogram of each child was monitored. The overall results of 63 visits indicated that 87-90% of the 235 desaturation episodes were due to patient movement. The clinical implications of the findings are discussed.