During the Carbonaceous Species Methods Comparison Study at Glendora, CA, six groups made independent measurements of ambient formaldehyde concentrations during the period August 11–21, 1986. Measurement methods included DNPH-impregnated cartridges, an enzymatic technique, a diffusion scrubber, Fourier transform infrared spectroscopy (FTIR), differential optical absorption spectroscopy (DOAS), and tunable diode laser absorption spectroscopy (TDLAS). Sufficient data were obtained over the 10-day period to assess differences among methods based on hourly averages, 4- and 8-hour time-integrated sampling periods, and 3–5-minute averaging times. Comparison among the three spectroscopic methods (DOAS, FTIR, and TDLAS) showed good agreement, within 15% of the mean of the three methods for 162 hourly values. The enzymatic technique and diffusion scrubber reported concentrations ∼ 25% higher and 25% lower than the spectroscopic mean, respectively, for the entire study period. The DNPH cartridges, the only routine monitoring method in the study, yielded values 15–20% lower than the spectroscopic mean, with somewhat lower values over longer sampling periods. Measurement response time was not a factor for either the FTIR or the TDLAS methods. Formaldehyde exhibited a morning and an afternoon peak each day during the study. The morning peak was shown to be associated with NOx and particulate black carbon concentrations; the afternoon peak was associated with the arrival of photochemically produced ozone at the site. The maximum peak hourly average, reported as the mean of the three spectroscopic methods, was 20 ppb; the average nighttime minimum was 6 ppb. These data suggest that primary formaldehyde may be a greater contributor to total formaldehyde levels than previously thought.