One hundred and forty-three sequential, unselected patients with rheumatoid arthritis (RA) were prospectively evaluated for evidence of secondary Sjögren's syndrome (sSS). All of them had a labial salivary gland (LSG) biopsy, since a greater than or equal to 2+ score (Tarpley's classification) was considered mandatory for the diagnosis of sSS. One hundred eleven were completely investigated. The diagnosis of sSS was established when a patient with a positive LSG biopsy (greater than or equal to 2+ score) had keratoconjunctivitis sicca (KCS) (positive rose bengal test and/or the combination of subjective xerophthalmia and positive Schirmer's I) and/or xerostomia (subjective xerostomia and decreased parotid flow rate = [PFR]). Forty-four patients had an LSG biopsy score of greater than or equal to 2+, 28 a 1+ score and 39 a negative biopsy. Thirty-four of the 44 patients with positive biopsy had sSS, suggesting a 31% prevalence of sSS in RA. The 1+ score group represents a nonspecific category, resembling the negative biopsy group. Rose bengal test was very sensitive for sSS, whereas Schirmer's I and PFR were not. Parotid gland enlargement was unusual, and extraglandular manifestations, with the exception of diffuse interstitial lung disease, were relatively uncommon in all biopsy groups. No patient complained of xerophthalmia or xerostomia on his/her own. Anti-Ro (SSA) antibodies, detected in 23.5% of the patients with sSS, correlated well with positive LSG biopsy. Our study suggests that sSS in RA is common, benign and subclinical, requiring a positive lip biopsy and specific tests for its diagnosis.