There are new speech recoginition systems in the market.These are called Continous speech recoginition systems. CSRs in the market today,by Dragon Systems, IBM, Lernout & Hauspie, and Philips are suboptimal.They don't work well enough yet. But all are making aggressive progress,they have not arrived yet.CSR are rapidly becoming adequate for Radiology and Pathology.But for other clinical areas serious limitations are seen. The scope and quality of clinical voabularies is improving gradually.Radiology and Pathology are two specialties where CSR when implemented require the least support of transcription. CSR allows Drs to have their dictated information encoded automatically into healthcare organization's system for immediate access. For the vast amount of documantation needs,CSR will bypass dictation-transcription cycle. In future CSR will provide data capture method for computer based record(CBR),or electronic chart.Transcriptionists will be there for quality assurance, checking, and correcting errors. Some of the factors which at present are preventing clinician acceptance are : 1.Increased errors caused by accents,pronunciation,etc. 2.Errors due to specialized clinical conditions and new terms 3.Inability of CSR to complete documantation with missing information. For this MTs serve as human check. 4.CSR are unable to ensure correct placement and sequencing of information. I don't think these problems can be rectified that easily. Not at least in near future.Till then MTs will continue in health care with same level of participation.Advantages of having MTs do the transcription are unavoidable. But also on other hand, we should keep in mind the expected role of such systems in the future and prepare ourselves for that.