The Modified Ashworth Scale (MAS) is commonly used in clinical practice for grading spasticity. However, it was modified recently by omitting grade "1+" of the MAS and redefining grade "2". The aim of this study was to investigate the inter-rater reliability of MAS and modified MAS (MMAS) for the assessment of poststroke elbow flexor spasticity. Sixty-four patients with poststroke hemiplegia were enrolled. The mean age and time since the stroke were 60.5 plus or minus 11.9 years and 15.7 plus or minus 10.2 weeks, respectively. The patients were tested by two raters having equal experience in applying MAS. They were not subjected to any training for administering MMAS. After performing no more than two test movements, the raters graded the resistance felt, according to the MAS and MMAS, respectively based on same stretch. The degree of agreement was analyzed using the weighted kappa (kw) statistic. Inter-rater agreements were very good for both MAS and MMAS, with weighted kappa values of 0.868 and 0.892, respectively. The highest agreements were observed for grade "0" in applying MAS and for grade "2" in applying MMAS; on the other hand, the lowest agreements were observed for grade "2" in applying MAS and for grade "3" in applying MMAS. According to our results, MAS and MMAS have very good inter-rater reliability for assessment of poststroke elbow flexor spasticity. Neither of the scales is superior to each other when using them to grade spasticity in patients with hemiplegia for this muscle group.