Workplace-accident statistics can change during economic downturns because firms alter employment, activity, work organization, and reporting, yet firm-level evidence often conflates differences between firms with changes occurring within the same firm. This longitudinal study examined associations of net sales and total assets with recorded workplace-accident incidence in 150 Spanish firms observed annually from 2006 to 2009. The balanced firm-year structure was analyzed using firm and calendar-year fixed effects with standard errors clustered by firm. Adjusted within-between decompositions distinguished within-firm changes from between-firm differences. The archived sample included firms that obtained OHSAS 18001 certification at different times and a non-certified comparison group; certification was therefore treated as a time-varying contextual characteristic rather than a uniform sample attribute. Higher average sales and assets characterized firms with lower average accident incidence, but changes in either economic measure within a firm were not associated with corresponding changes in incidence after accounting for workforce and common calendar-year conditions. Recorded incidence was lower in 2009 than in 2006, although this calendar-year contrast cannot identify a causal recession effect. Results were consistent across alternative outcome definitions, complete-sample restrictions, sector-specific time trends, outlier checks, first-difference models, and alternative outcome transformations. The findings show why between-firm and within-firm evidence should be separated when interpreting occupational-injury patterns during macroeconomic disruption. The longitudinal estimates therefore do not support the hypothesis that improvements in firm economic conditions were associated with reductions in accident incidence within firms.