Stroke is the leading cause of disability in developed countries, and the major cause of mortality in Brazil.It is associated with low quality of life, disability, decreased muscle strength and control due to the hypertonia, spasticity. These pathways can develop inadequate upper and lower limbs functional mechanisms. These changes are linked with disabilities and painful syndromes.

The shoulder pain relationed after stroke has a variable prevalence from 16 to 84%, from mild to severe pain and is relationed as a precursor factor of secondary deformities, depression, longer hospital stay. Its etiology remains controversial with many possibilities as rotator cuff injuries, glenohumeral dislocation, impact syndrome, bicipital tendinitis, hand shoulder syndrome, myofascial painful syndrome, presence of spasticity and contractures, adhesive capsulitis, central pain and others. Its management is controversial and could be done with physical therapy (kinesiotherapy) intra-articular or local injections (muscles and nerves), functional electrical stimulation, acupuncture, herbal medicine, tapping, myofascial painful syndrome treatment, painkillers, anti-inflammatories and antispastic drugs like botulinum toxin A - tested in few studies, case series, against corticosteroids.

The aim of this study will be to evaluate the effectiveness of the use of 200 units of abobotulinumtoxin against placebo in both pectoralis major and subscapularis: for reducing shoulder pain (Visual Analogue Scale, McGill pain scale), active and passive range of motion(goniometer),upper limb function (Fugl-Meyer test), burden of care questionnaire (apply to caregivers).

This study is designed as a prospective, double-blind, randomized, controlled study in two Rehabilitation Centers (Rehabilitation Center from Santa Casa de São Paulo and Hospital de Clínicas of Universidade de Sao Paulo - Ribeirão Preto).

The principal aim will be the pain evaluation (Visual Analogue Scale and McGill pain scale) with the use of 200 units of abobotulinumtoxin (Dysport) against placebo in both pectoralis major and subscapularis, after 01and 04 months of the procedure The secondary objectives will be analysis the active and passive affected shoulder range of motion (goniometer), the upper limb function (Fugl-Meyer test), Burden of care questionnaire (apply to caregivers) and the use of medicine and therapies.

The sample size was calculated to a alpha error (5%), statistical power (80%) and a decrease of 1.4 points in Visual Analogue Scale, totalizing 10 patients per group, to avoid lack of statistical power due to possible drop out we will recruit 12 patients per group.

The muscles selected for this study are the pectoralis major and subscapularis muscles, each receiving 200U of Dysport® distributed in 2 points, guided by electrical stimulation .The evaluations will be performed in 0, 1 and 4 months after the procedure.

The researchers that will apply and evaluate the patients will be blinded. The patients will be randomized by 6 blocks of four.

The inclusion criteria will be spasticity in upper limb due to ischemic or hemorrhagic hemispheric stroke; over 18 years old; diagnosis of hemiplegic painful shoulder syndrome, regardless of motor dominance; agreement of the patient, family member and / or responsible caregiver to participate in the study.

The exclusion criteria will be pain before stroke in shoulder affected by hemiplegia; previous treatment using TXB-A for painful shoulder; cognitive impairment that hinders assessment and collaboration with treatment; counterindication to the botulinum toxin use; structured joint deformity in the shoulder affected by pain.

The qualitative variables will be described through percentages of each categories. The quantitative variables will be analyzed for normality by the Komolgorov-Smirnov. The comparison of the value average of the pain variable evaluated by VAS will be performed by the ANOVA test, considering the treatment groups and the moment of evaluation, followed by post hoc comparisons, if the ANOVA test indicates a significant effect group time.The dropouts will be evaluated according to the intention to treat. We will use software Stata11 or similar model

Inclusion Criteria:

• Spasticity in upper limb due to ischemic or hemorrhagic hemispheric stroke;
• Diagnosis of hemiplegic painful shoulder syndrome, regardless of motor dominance;
• Agreement of the patient, family member and / or responsible caregiver to participate in the study.

Exclusion Criteria:

• Pain before stroke in shoulder affected by hemiplegia;
• Previous treatment using TXB-A for painful shoulder;
• Cognitive impairment that hinders assessment and collaboration with treatment;
• Counter indication to the botulinum toxin use; Structured joint deformity in the shoulder affected by pain

The patients will receive placebo or botulinum toxin in the same characteristics. Just a assessor of the study will know the groups