I have been some time ago asked to give a lecture to DAMTP research students on the theme of "How to write a paper?". Realising that the subject matter of the talk might be of wider interest, I am happy to provide it here as a pdf file.

Few words of caution, though:

• When I am saying "paper", I mean "mathematical paper". Please, don't use it as a guide to writing papers in ontology, oncology or palaenotology.
• This is my personal take on an issue that admits endless personal takes. Caveat emptor.
• The slides of the talk have been accompanied by 60 minutes of narrative which, needless to say, is missing from the above file.
• You might have entirely different ideas on how to write a paper or on what I have missed or got wrong in my presentation. You might even be right. Yet, ideally, I would prefer not to enter into endless correspondence on such matters.

Quasi-legal rejoinder:
Following (or appearing to follow in a subjective manner) my suggestions, prescriptions, recipes or other ideas, as presented in the pdf file above, whether understood or misunderstood, is done entirely at your own risk. I thereby do not undertake or imply or promise or give you a shadow of an illusion of a vague promise that your paper will be accepted, praised, looked upon indulgently, tolerated or otherwise approved by any managing editor, editor, referee or reader. And this includes me in all four of these categories.