Evaluate the intra- and inter-rater reliability of maximal voluntary contraction (MVC) of pelvic floor muscle (PFM) assessment by unidigital and bidigital vaginal palpation. As a secondary aim, evaluate the construct validity of vaginal palpation and a manometer.A total of 120 women were recruited and allocated into three groups according to age (group 1 [G1]: 18-35; G2: 36-59; G3: ≥60 years). Examiner A conducted an interview and examiners B and C performed evaluations of PFM function using both unidigital and bidigital vaginal palpation. Seven to 10 days later, examiner B repeated both types of vaginal palpation and measured the MVC pressure with a Peritron. Reliability was analyzed using Cohen's linear weighted Kappa (κ ). The construct validity was calculated using the Spearman's correlation (r ).Intra-rater reliability of unidigital and bidigital palpation was κ  = 0.75 and κ  = 0.58 in G1, κ  = 0.59 and κ  = 0.73 in G2, and κ  = 0.79 and κ  = 0.86 in G3, respectively. Inter-rater reliability of unidigital and bidigital palpation was κ  = 0.52 and κ  = 0.48 in G1, κ  = 0.47 and κ  = 0.52 in G2, and κ  = 0.50 and κ  = 0.64 in G3, respectively. Spearman's correlation coefficients were significant at r  = 0.79 and r  = 0.80 for unidigital and bidigital vaginal palpation, respectively.Unidigital or bidigital vaginal palpation were indicated to evaluate G1 and G2 but bidigital palpation was more reliable in G3 when two examiners performed the evaluation. For the assessment conducted by one examiner, unidigital palpation was more reliable in G1 while bidigital palpation was more reliable in G2 and G3. Both unidigital and bidigital palpation have high measures of validity.