During 1976-1978, there was an overall decline of arson in Seattle. A spatial examination of the city by sections shows that there appears to be an upsurge of arson activity in the proximity of downtown Seattle, and an overall improvement in other urban locales. However, spatial data alone cannot explain the reasons or causes of why certain areas did better or worse over the 1976-78 period. An evaluation of block areas within census tracts might help to explain that data. An examination of arson by building occupancy categories (residential, mercantile, storage, office, public assembly, vacant, industrial, educational, and institutional) and by the type of structure reveals that: (1) residential arson fires rose dramatically in 1976 in both number and in dollar loss. During 1977 and 1978 residential arson rose in greater proportion to all arson during the 3-year period; (2) considerable improvement was experienced in the volume and losses in mercantile and vacant building categories, as well as educational structures; (3) considerable variability occurred in many occupancy classifications; and (4) a significant number (32 percent) of arson fires reported no dollar loss due to flame damage, and a majority (59.6 percent) of arson cases reported damage below $100. An adjustment using the Comparative Cost Multipliers is necessary in order to compare the fire loss statistics. Once that adjustment is made, it is seen that most of the categories fluctuate widely. Except for residential loss, which remains the largest single category of fire loss, none of the categories even remain in the same order of magnitude from 1967-1978. Six tables illustrate the text.